>>> [snip] Is there any ways of cancelling a search command from a query >>> transformer? >> >> No. [snip] > > Could it then be an idea to standarize a way of doing this? E.g. > letting the query transformer set a property of some sort on the > search command, that (all of) the search commands read before they > actually do anything?
I cant think of any elegant way of doing so. But tying it into another general problem: the way query transformers use the junkyard to notify the search command of altered state; maybe there's a way to have general search command properties that the query transformer can set/update. These properties would become part of the ResultList when it is time to construct it. Can you integrate this and provide suggestions with the new SearchCommand design proposal? ~mck _______________________________________________ Kernel-development mailing list [email protected] http://sesat.no/mailman/listinfo/kernel-development
